Skip to main content
. Author manuscript; available in PMC: 2023 Nov 21.
Published in final edited form as: Mathematics (Basel). 2023 Jan 20;11(3):557. doi: 10.3390/math11030557

Figure 1.

Figure 1.

Distribution of λopt (A) across 1000 runs of glmnet is shown, along with the distribution of the corresponding number of predictors in the signatures (B). Each application of the glmnet algorithm for this dataset results in an optimal signature, with size varying greatly from less than 5 to over 50 predictors.